ray diffraction

The inclusion complexes were characterized by an X-ray powder diffractometer (D8 Advance diffractometer, Bruker AXS GmbH, Billerica, MA, USA) using Cu K-α

radiation (λ = 1.5406 Å). Each sample was scanned at 40 kV and 40 mA in the interval 3-40° 2 ϴ (at 0.1/s scanning speed and with 0.01° step size). Prednisolone, MAβCD and PASP-SH10-CD1 were measured as solid samples without further sample preparation.

MAβCD -prednisolone and PASP-SH10-CD1-prednisolone complexes were measured as the freeze-dried supernatant of the suspensions prepared for the phase solubility study containing 2.5 wt% MAβCD and 10 wt% PASP-SH10-CD1, respectively.

Rheology of gelation

Gelation of the aqueous solutions of the polymers was monitored by oscillation rheometry using an Anton Paar Physica MCR101 rheometer (Austria) with plate–plate geometry (sample gap: 0.100 mm). Three compositions were analyzed: PASP-SH10, PASP-SH10-CD1 without prednisolone, and with prednisolone (PASP-SH10-CD1-PR).

Precursor polymer solutions were prepared by dissolving 15 mg polymer in 105 µl of PBS. In the case of PASP-SH10-CD1-PR 0.2 mg prednisolone was also dispersed in the PBS. The precursor polymer solutions were mixed with 30 µl 1 M NaBrO3 solution on the plate of the rheometer to initiate gelation. Storage modulus was monitored in oscillatory mode at constant strain and angular frequency (γ = 1%, ω = 10 rad/s). After 20 min of gelation, the frequency dependence of the storage moduli (γ = 1%, ω = 1-100 rad/s) of the resultant gels were also recorded.

Release of prednisolone

The drug diffusion profile of prednisolone was determined with a vertical Franz diffusion cell system (Hanson Microette Plus™). As a donor phase 30 mg of PASP-SH10-CD1 was dissolved and 0.3 mg prednisolone was dispersed in 210 µl PBS. Then, 60 µl 1 M NaBrO3 solution was added to initiate gelation and the mixture was placed on Porafilm diffusion membrane (pore size of 0.45 µm, previously impregnated with PBS).

As a reference formulation the mixture of unbound MAβCD and PASP-SH10 was used.

In this case 30 mg PASP-SH10 and 2 mg MAβCD was dissolved instead of PASP-SH10-CD1 while the other components remained the same. As a second reference formulation, 300 µl of 0.1 wt% prednisolone suspension in PBS was used. The acceptor phase was 7.0 ml PBS and it was thermostated at the ocular surface temperature, 35 °C. The duration of the measurement was 24 h and three parallel measurements were carried out. Samples of 0.8 ml were taken from the acceptor phase at different times by the autosampler and replaced with fresh acceptor phase. The prednisolone released was measured at 247 nm by a UV-Vis spectrophotometer (Thermo Scientific Evolution 201).

Results and discussion

Synthesis and chemical characterization

Synthesis of cyclodextrin-modified thiolated poly(aspartic acid) (PASP-SH10-CD1) was performed in a three-step reaction. First, PSI was modified with 6-deoxy-6-monoamino-β-cyclodextrin hydrochloride in the presence of dibutylamine in order to

Enhancing the solubility of prednisolone…

deprotonate the amine group. In the second step, thiolation was performed by the addition of cysteamine under nitrogen atmosphere. Finally, the unreacted succinimide rings were partially hydrolyzed to aspartic acid residues in a mildly alkaline buffer solution. The chemical reaction is shown in Figure 4.1.

Figure 4.1 Synthesis of cyclodextrin-modified thiolated poly(aspartic acid).

The chemical structure was characterized with ¹H NMR spectroscopy (Figure 4.2).

The peaks at 4.63 and 4.46 ppm represent the methyne protons (nuclei 1) of the aspartic acid and modified aspartic acid repeating units, and the peak at 2.74 ppm is assigned to their methylene protons (nuclei 2). The methyne and methylene protons (nuclei 5 and 6) of the unreacted succinimide rings are at 4.92 and 3.14 ppm, respectively. Nuclei 6 gives a peak also at ca. 2.80 ppm that overlaps with the peak of nuclei 2 [6]. The methylene protons of cysteamine side groups are at 3.34 and 2.63 ppm (nuclei 3 and 4, respectively).

The protons of the cyclodextrin rings being in an O-CH-O bond (nuclei B, 7 protons in each ring) give a peak at 5.04 ppm, and all other cyclodextrin protons (nuclei A, 42 protons in each ring) have a chemical shift in the range of 3.5-4.0 ppm. The chemical shifts are summarized in Table 4.1.

Table 4.1 ¹H NMR peak assignment of PASP-SH10-CD1.

Proton number

Chemical shift (ppm)

1 4.63, 4.46

2 2.74

3 3.34

4 2.63

5 4.92

6 3.14, 2.80

A 3.5-4.0

B 5.04

Figure 4.2 (a) ¹H NMR spectrum and (b) expected chemical structure of PASP-SH10-CD1.

The ratio of the different repeating units was calculated based on the integral intensities of the characteristic NMR peaks. The degree of thiolation (XSH-NMR) was calculated by finding the ratio of the thiolated repeating units and the total number of repeating units according to Eq. 4.1.

𝑿_{𝑺𝑯−𝑵𝑴𝑹}= 𝐴_{3.34 𝑝𝑝𝑚}

𝐴2.5−3.0 ppm− 𝐴_{3.34 𝑝𝑝𝑚}+ 𝐴_{3.14 𝑝𝑝𝑚}

This formula is analogous to Eq 3.2; the thiolated repeating units are represented by the integrated intensity of nuclei 3. In the denominator, nuclei 2 represents all modified and hydrolyzed repeating units and to calculate its area, the area of nuclei 4 (which equals

(4.1)

a) b)

Enhancing the solubility of prednisolone…

to nuclei 3, i.e. A3.34 ppm,) must be subtracted from the total area integrated between 2.3-3.0 ppm (A2.5-3.0 ppm). The methylene peak area (nuclei 6) of unreacted succinimide rings at 3.14 ppm also needs to be added to the denominator (its other peak at ca. 2.80 ppm is already included in A2.5-3.0 ppm). The peak of oxidized side groups at 3.51 ppm (in comparison with Eq 3.1 and 3.2) is not included in the denominator due to peak overlapping with the peaks of β-CD. According to the spectrum, the amount of oxidized side groups is rather low, and based on the earlier calculations omitting those leads to an acceptable error.

The number of cyclodextrin rings compared to the total number of repeating units (XCD-NMR) can be calculated by Eq. 4.2. Areas are normalized because A3.5-4.0 ppm

represents 42 protons and the peaks in the denominator represent 2 protons each.

𝑿𝑪𝑫−𝑵𝑴𝑹= 𝐴3.5−4.0 ppm/42

(𝐴2.5−3.0 ppm− 𝐴3.34 𝑝𝑝𝑚+ 𝐴3.14 𝑝𝑝𝑚)/2

And the ratio of residual succinimide rings (XSu-NMR) was calculated with Eq. 4.3. In order to calculate the total area representing the unreacted succinimides, A3.14ppm must be multiplied by two due to the other peak of nucleus 6 ca. at 2.80 ppm (equality of their areas is assumed):

𝑿𝑺𝒖−𝑵𝑴𝑹 = 2 ∙ 𝐴_{3.14 𝑝𝑝𝑚}

𝐴2.5−3.0 ppm− 𝐴3.34 𝑝𝑝𝑚+ 𝐴3.14 𝑝𝑝𝑚

XSH-NMR was found to be 8.0% and XCD-NMR was 0.75% The ratio of residual succinimide rings (XSu-NMR) was 35%.

Unlike in the case of PASP-SH polymers described in Chapter 3, the alkaline hydrolysis of the polymer was not complete, and a rather large amount of residual succinimide rings remained in the product. The reasons for this are the milder alkaline hydrolysis that was applied (pH = 8 instead of pH = 13) in order to preserve the stability of cyclodextrin rings. In spite of the lower conversion of the hydrolysis, the solubility of the PASP-SH10-CD1 remained similar to that of PASP-SH10 in the concentration range used.

Phase solubility

The phase solubility curves of the inclusion complexes of prednisolone formed with MAβCD and PASP-SH10-CD1 are plotted in Figure 4.3. The phase solubility profile is linear in both cases (Type A profile according to the classification of Higuchi and Connors [7]) which is common in the case of water soluble drug-CD complexes.

(4.2)

(4.3)

0 5 10 15 20 25 0

3 6 9 12 15

MACD

Prednisolone (mM)

Equivalent cyclodextrin (mM) PASP-SH10-CD1

Figure 4.3 Phase solubility curve of prednisolone with MAβCD and PASP-SH10-CD1.

The stability constant of the complexation (Kc) can calculated with Eq 1.14. The slopes of the fitted linear curves (a) and the Kc values are summarized in Table 4.1 Table 4.1 Slopes (a) and stability constants (Kc) of the phase-solubility study

Sample a Kc (1/M)

MAβCD 0.5696 1728

PASP-SH10-CD1 0.5203 1416

The optimal range of Kc for the solubilization of lipophilic drugs is 200-5000 1/M proposed by Szejtli [8] as in case of low values, solubilization is insufficient, while in case of higher values the absorption of the drug is limited. In our case both the small molecule and the modified polymer have a Kc values in this favorable range. The fact that the Kc values are rather close to each other indicated that the complexation ability of MAβCD changed only slightly by its chemical immobilization onto the polymer.